Usability of affective interfaces for a digital arts tutoring system

Affective computing techniques have become increasingly important as advanced education technologies. By applying these techniques to education, this work designs and evaluates a novel Affective Tutoring System for the Digital Arts (ATSDAs). By semantically analysing a text with ontological references, the emotions induced by a text when input by a user are identified. Inference of emotions is accomplished using OMCSNet and WordNet, two engines commonly used in computational linguistics research. The proposed system has a visual agent that provides text feedback based on inferred emotions from textual analysis. The proposed system has a conscientious design flow that includes concept modelling, prototype design, expert-based evaluation (which consists of a cognitive walkthrough and heuristic evaluation), final system design and a series of evaluations from a learner's perspective. The System Usability Scale (SUS) evaluation results show that this system achieves positive usability and learners enjoy interacting with the proposed system.